WHAT IS ASTRAGALUS HYPOGLOTTIS ? 227 



A. epiglottis, to both of which it is applied in the Mantissa? 

 None of the three are erect plants. The stems in all are ascending ; 

 more or less prostrate or erect according to whether the plants 

 grow on bare ground, or among herbage. Only last June I found 

 both purpiireus and danicus plentifully in the neighbourhood of 

 Gap, in Dauphine. In both species the stems lie about among 

 the grass, and only the flowering-heads show well up above it. I 

 think this objection may be dismissed as cavillous. 



(3) Flowers 8-10. No doubt the flowers in each head of 

 A. purpureus are normally more numerous. But in the specimen 

 in Herb. Linn., as already mentioned, one of the heads contains 

 about nine flowers, and the other only about five. So that if 

 Linnaeus described from the plant actually under his eyes the 

 number he assigns would be quite correct. This objection also 

 may therefore be dismissed. 



(4) " Gousse recourbee en crochet " ; in the Danish version, 

 " den i Spidsen hagekrummede og toklovede Baelge." Surely 

 this is not a fair translation of Linnasus's words, " Legumina 

 replicata . . . acumine . . . subulato recurvo," and " leguminibus 

 replicatis . . . acumine reflexo." There is not a word that justifies 

 "en crochet." It is not the legumes that are "recurva" or 

 "reflexa," but their styles, their "acumen." The legumes them- 

 selves are "replicata," which does not mean reflexed or recurved, 

 but laterally folded in (more or less) upon themselves. It is 

 equivalent to " complicata," only outwards, not inwards. This is 

 clear from the very same word being used of the legumes of 

 epiglottis and pentaglottis. Linnaeus has paraphrased the same 

 idea in the expressions "lateribus conniventibus " and "extrorsum 

 conduplicata " applied to those other two species respectively. 

 For the legumes of epiglottis, which, instead of being erect, are 

 pendulous, he has employed the words "cernuis" and "pendulis." 

 In those of A. purpureus, the base of the long style is straight, 

 but the upper part of it is flexible, curved, and slightly hooked, in 

 fact, " acumine subulato recurvo." 



Smith and Sibthorp saw no difficulty in accepting this account 

 of the legume, both for the British species and for the Linnean 

 specimen of A. piuyureus, so that it can hardly be an impossible 

 description of the latter. 



(5) " Bifurcated at the summit," " gousse bifurquee a son 

 sommet." But here again it is only the style, not the body of the 

 legume, which Linnaeus declares to be double when ripe ; "acumine 

 duplici in maturis." I have only been able to see very few 

 specimens of purpureus with dead ripe legumes, and they seem 

 to have a sort of double beak, though less conspicuously so than 

 in epiglottis. 



Of these five objections the first and the fifth are the only ones 

 that make any serious difficulty. Surely they are entirely out- 

 weighed by the words " Similitudinem gerit cum A. pentaglotto," 

 which suits purpureus better than any other species, and by the 

 statement that the legumes are "tecta pihs (non furfuraceis) longis 

 albis moUibus." This will only suit purpureus or danicus, and 



R 2 



